N 7-methylguanosine-related gene risk signature as a potential prognostic biomarker in hepatocellular carcinoma

doi:10.21203/rs.3.rs-1601394/v1

Download PDF

Research Article

N 7-methylguanosine-related gene risk signature as a potential prognostic biomarker in hepatocellular carcinoma

https://doi.org/10.21203/rs.3.rs-1601394/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background: Hepatocellular carcinoma (HCC) remains one of the most common malignancies worldwide, with an increasing annual incidence and mortality. N 7-methylguanosine (m7G) is a modification in which a methyl group is added to the seventh N position of messenger ribonucleic acid (RNA) guanine (G) by methylation transferase action. Increasing evidence revealed that m7G-related genes are closely associated with HCC development. This study aimed to develop a new m7G-related prognostic model to assess the prognosis of patients with HCC.

Methods: Herein, we used RNA transcriptome sequencing and clinical data of patients with HCC from the Cancer Genome Atlas (TCGA). Patients with HCC were classified into two subtypes based on the m7G gene, and differential expression genes (DEGs) between the two types were identified. Prognostic risk models were constructed by assessing the prognostic significance of these DEGs using univariate Cox, least absolute shrinkage and selection operator, and multivariate Cox regression analyses. Functional differences were compared between the high and low-risk groups and differences in immune infiltration and immunotherapy. Finally, a preliminary pan-cancer analysis of the two genes in the model was performed.

Results: Significant survival differences were found between different subtypes of patients with HCC, who were divided into high and low-risk groups according to the median risk scores. The Kaplan–Meier survival analysis showed a substantially better overall survival in the low-risk group than that in the high-risk group (p < 0.001). Receiver operating characteristic curves and within-group validation further confirmed the model’s reliability and accuracy. Additionally, significant differences were found in functional analysis, immune infiltration, and immunotherapy effects between the two groups. GAS2L3 and HTR1D have prognostic value in various cancers, which may help for subsequent studies.

Conclusions: In conclusion, this prognostic model based on m7G-related genes is beneficial for individualized prognosis assessment of patients with HCC and provides a reference for clinical treatment.

m7G

Genes

Hepatocellular carcinoma

Prognostic signature

Bioinformatics

Liver cancer remains one of the most common malignant tumors, ranking seventh for incidence and second for mortality worldwide according to the latest data released by the International Agency for Research on Cancer in 2020¹. Of the liver cancers, > 80% are hepatocellular carcinoma, which causes a serious economic burden as common cancer in most countries in the world². HCC development is closely related to several factors, such as viral hepatitis and cirrhosis, fatty liver, and alcohol³, and current treatments include hepatectomy, transcatheter arterial chemoembolization, drugs, liver transplantation, etc⁴. Current therapeutic approaches have prolonged the survival time of patients with HCC to some extent; however, the prognosis remains a considerable challenge in clinical work due to the high heterogeneity of HCC. Therefore, discovering novel and reliable methods to assess the prognosis of patients is urgently needed, as well as prompt management of patients ^5,6.

More and more scholars focus on the way of post-transcriptional modifications to ribonucleic acid by exploring cancer development from the transcriptional level perspective. RNA can alter its role in translation and splicing through > 170 different types of modifications, commonly including N 6-methyladenosine (m6A), 5-methylcytidine (m5C), N 1-methyladenosine (m1A), N 4 acetylcytidine (ac4C), ribose methylations (Nm), and N 7-methylguanosine (m7G)⁷. Previous studies have helped us understand m6A, m5C, etc, but the role of m7G in HCC using bioinformatics has not been reported, which deserves further investigation. M7G is a modification in which a methyl group is added to the seventh N position of messenger RNA guanine (G) by methylation transferase action⁸. M7G modification is one of the most common forms of base modification in post-transcriptional regulation and is widely distributed in tRNA, rRNA, and eukaryotic mRNA of many organisms⁹. Additionally, a few archaeal has been reported to confirm the occurrence while its exact location is not well understood. In tRNA, m7G often occurs at position 46 of its variable region, which generates charge at specific tRNA sites by methylation and facilitates the stability of tRNA three-dimensional structure. Studies have shown that m7G modification of tRNA may be an important factor in fungal infection of hosts and its close linkage with certain genetic diseases^10,11. The m7G cap plays an important regulatory role for various biological functions of mRNA, is required for mRNA splicing, and initiates translation of mRNA in the cytoplasm^12,13. m7G modification occurs at position G1405 of 16SrRNA, which gives it aminoglycosidic activity¹⁴. In conclusion, as one of the important RNA modification modalities, m7G has an important role in maintaining RNA processing metabolism, stability, exonucleation, and protein translation, but has not received sufficient attention, thus we conducted this study.

In this study, we established a risk signature related to m7G-associated genes with strong and powerful predictions by analyzing genome-wide data of patients with HCC from TCGA database, which has never been reported in previous studies. All patients with HCC were classified into two subtypes based on m7G-associated genes, and a significant survival difference between these two types was demonstrated. Univariate Cox, least absolute shrinkage and selection operator (LASSO)-Cox, and multivariate regression analyses were used to screen out three m7g-associated genes (GAS2L3, HTR1D, and NEIL3) to construct a prognostic signature. Kaplan–Meier plots, time-dependent receiver operating characteristic (ROC), risk curve, and scatter plot analyses were performed to validate the accuracy of model predictions. Considering that m7G is closely associated with tumor immunity^15,16, differences in immune characteristics immunotherapeutic evasion between high- and low-risk groups were further analyzed. Additionally, a pan-cancer analysis of two genes (GAS2L2 and HTR1D) in the risk model was performed.

Data collection

Gene expressions of transcriptome profiles and clinical data of 33 types of cancer were obtained from TCGA (https://tcga-data.nci.nih.gov/tcga), including adrenocortical carcinoma (ACC), bladder urothelial carcinoma (BLCA), breast invasive carcinoma (BRCA), cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC), cholangiocarcinoma (CHOL), colon adenocarcinoma (COAD), lymphoid neoplasm diffuse large b-cell lymphoma (DLBC), esophageal carcinoma (ESCA), glioblastoma multiforme (GBM), head and neck squamous cell carcinoma (HNSC), kidney chromophobe (KICH), kidney renal clear cell carcinoma (KIRC), kidney renal papillary cell carcinoma (KIRP), acute myeloid leukemia (LAML), brain lower grade glioma (LGG), liver hepatocellular carcinoma (LIHC), lung adenocarcinoma (LUAD), lung squamous cell carcinoma, mesothelioma (MESO), ovarian serous cystadenocarcinoma (OV), pancreatic adenocarcinoma (PAAD), pheochromocytoma and paraganglioma (PCPG), prostate adenocarcinoma (PRAD), rectum adenocarcinoma (READ), Sarcoma (SARC), skin cutaneous melanoma (SKCM), stomach adenocarcinoma (STAD), testicular germ cell tumors (TGCT), thyroid carcinoma (THCA), thymoma (THYM), uterine corpus endometrial carcinoma (UCEC), uveal melanoma (UVM), and uterine carcinosarcoma. The twenty-nine m7G genes were extracted from previous published literatures including METTL1, WDR4, NSUN2, DCP2, DCPS, NUDT10, NUDT11, NUDT16, NUDT3, NUDT4, NUDT4B, AGO2, CYFIP1, EIF4E, EIF4E1B, EIF4E2, EIF4E3, GEMIN5, LARP1, NCBP1, NCBP2, NCBP3, EIF3D, EIF4A1, EIF4G3, IFIT5, LSM1, NCBP2L, and SNUPN.

Consensus clustering based on m7G genes

First, the m7G gene expression in HCC samples were extracted using the “limma” package in R software. Then, based on the m7G gene expression, the “ConsensusClusterPlus” package in R software was used to perform consensus clustering analysis of all patients with HCC. Patients with HCC were divide into k (k = 2–9) clusters and then survival analyses was done from these k clusters to see whether the clustering is reasonable. Afterward, the differentially expressed genes between different clusters were identified by the “limma” package in R software. The intersecting DEGs were extracted for next analysis.

Building a prognostic model to predict the overall survival of all patients with HCC

First, the prognostic value of DEGs was obtained using univariate Cox regression analysis. Second, these prognostic DEGs with P-values of <0.05 in the univariate analysis were selected for a LASSO regression analysis and the results were included into a multivariate Cox model to construct a risk signature. The risk score = Ʃ (βi × EXPi), where βi (coefficients) represented the weight of the respective gene and EXPi represented the expression level. All patients with HCC were divided into high-risk (over the median risk score) and low-risk groups (less than or equal to the median risk score) according to the median risk score. The two groups were used in a follow-up study.

Validation of the risk score

All the obtained HCC samples from TCGA were randomly assigned into training and internal validation cohorts at a 1:1 ratio to assess whether this prognostic model has the similar predictive power in different groups. The training and internal validation group were also divided into high-risk and low-risk groups according to the same cutoff value. Risk score was calculated for each patient with HCC using the above formula. Similarly, the two groups were subjected to KM, ROC curve, risk curve, and scatter plot analyses.

Independent prognostic and clinical correlation analysis

Univariate and multifactorial analyses of the risk score together with numerous clinicopathological conditions, including age, gender, grade, and T, N, and M stages, were performed to evaluate the use of the prognostic model as an independent factor to predict the prognosis of patients with HCC. Additionally, KM analysis was performed for different clinicopathological subgroups based on risk scores.

Construction of nomogram

We constructed a nomogram consisting of clinical parameters and risk score using “rms” package in R software. The area under the curve (AUC) of the standard ROC curve was plotted to compare the predictive abilities of the nomogram with other prognostic factors. The calibration chart is constructed to evaluate the consistency between the predicted survival probability and the actual value.

Functional enrichment analysis

Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses were performed using the “clusterProfiler” package in R software to clarify biological functions that the differentially expressed genes between high and low-risk groups are involved in. The differentially expressed genes between high and low-risk groups were extracted by “limma” package in R software.

Estimation of tumor-infiltrating immune cells and immunotherapeutic evasion with risk assessment signature

In the whole TCGA cohort, we analyzed the enrichment scores of immune cells and immune-related functions in the high and low-risk groups based on all patients with HCC using single-sample sequence set enrichment (ssGSEA). Additionally, the differences in TIDE, microsatellite instability (MSI), immune dysfunction, and immune exclusion scores between the two groups were analyzed based on the TIDE online platform because of a significant effect on immunotherapy evaluation.

GAS2L3 and HTR1D expression and survival analysis and relationship with clinical stage in cancers

A pan-cancer analysis of GAS2L3 and HTR1D was performed to assess their differential expression in tumor and normal tissues and their expression levels in different tissue types. Additionally, the correlation between GAS2L3 and HTR1D expression and cancer prognosis and clinical information were analyzed.

Statistical analysis

All statistical analyses were performed using R software (version 4.0.1). The Student’s t-test was applied to identify statistically differentially expressed genes. Cox regression and Kaplan–Meier curves with the log-rank test were used to evaluate the survival data. Fisher’s exact test was used to evaluate the significant GO terms and KEGG enrichment pathways. The ROC curve was used to determine the predictive accuracy of the risk signature. P-values of <0.05 were considered statistically significant.

Tumor classification based on the m7G genes

The overall workflow of the process that we used to develop and validate the risk signature to predict prognostic outcomes was depicted in Fig. 1. Based on the expression similarity of m7G genes, k = 2 seemed to be an adequate selection with clustering stability increasing from k = 2–9 in the TCGA datasets, which we named as cluster1 and cluster2 (Figures. 2A–2C). A significant difference of overall survival (OS) was noticed between these two clusters (p = 0.03) (Fig. 2D), and 273 m7G related differential expression genes (DEGs) (|log2FC| ≥ 1 and p < 0.05) were shown to be linked with these two clusters (Figure.2E). These results are believed to reveal a potential correlation of m7G related DEGs and patients with HCC, which led to the following exploration.

Constructing the prognostic model from m7G-related clustering

Firstly, 140 DEGs with prognostic value for patients with HCC were selected by univariate regression analysis in supplementary materials of Table 1 (p < 0.05). We selected 40 genes with the smallest p-values to analyze their mutation rates and types (Fig. 3A). The left panel shows the genes ordered by their mutation frequencies and the right panel presents different mutation types. Then, Lasso-Cox regression was performed to choose 12 genes, which were then included in multivariate Cox analysis (Fig. 3B and 3C). Finally, 3 genes (GAS2L3, HTR1D, and NEIL3) were identified as independent prognostic factors and selected for the prognostic model construction (Fig. 3D–F). The formula of the risk score for each patient was as follows: risk score =(1.18908*ExpHTR1D) + (2.31081*ExpGAS2L3) + (1.65707*ExpNEIL3). A follow-up analysis was conducted to assess the value of this prognostic model. All patients with HCC from TCGA databases were divided into two groups (low-risk and high-risk groups) based on the median risk score. The Kaplan–Meier curve indicated that high-risk patients had a substantially shorter OS than low-risk patients (p < 0.001, Fig. 4A). Time-dependent ROC curves were constructed to evaluate the prognostic model accuracy. As shown in Fig. 4B, the AUC values for 1, 3, and 5 years were 0.758, 0.698, and 0.644, respectively. From the scatter chart, we found that individuals at high-risk died sooner than those at low-risk (Fig. 4C and 4D). The PCA and t-SNE results showed that patients with varying risks were divided into two groups (Fig. 4E and 4F).

Verification of the m7G-related genes prognostic signature

The entire TCGA cohort (n = 370) was randomly divided into the training cohort (n = 186) and internal validation cohort (n = 184). In the training group (P = 007) (Figure.5A) and in validation group (P < 0.001) (Figure.5E), the ROC and Risk Score distribution point plots were also similar to the plots for the combined group, with significantly more patients dying in the high-risk group (Figure.5B–D, 5F–H). All these results suggested that these 3 m7G-related genes would be a valuable HCC prognostic model.

Independent prognostic value of the prognostic model

Risk score and clinicopathological information, such as age, gender, grade, and T, N, and M stages, were included for univariate and multifactorial analyses to assess the independent prognostic value of this prognostic model. The univariate cox analysis revealed that N stage (p < 0.001), T stage (p < 0.001), M stage (p < 0.021), and risk score (p < 0.001) were significantly associated with OS of patients with HCC (Fig. 4A). The multivariate cox regression revealed a risk score (p < 0.001) that was independent risk factors in the OS of patients with HCC (Fig. 4B).

Subgroup OS analysis with different clinicopathological features for all patients with HCC in TCGA cohort

Patients were stratified into subgroups based on distinct clinicopathological features, including age (age > 65 years and age ≤ 65 years), gender (female and male), grade (G1–G2 and G3–G4), clinical stage (stages I–II and III–IV) and T stage (T1–T2, T3–T4), to further assess the effectiveness of the prognostic model to predict the prognosis of patients with HCC in different clinicopathological conditions. Patients in each group were divided into high-risk and low-risk subdivisions based on the median risk score, and the groups were subjected to KM survival analysis. Results showed that the OS rate of the subgroups, including age > 65 (p = 0.002) and age ≤ 65 (p = 0.002) years, female (p = 0.039) and male (p < 0.001), G1–G2 (p < 0.001), G3–G4 (p = 0.041), stages I–II (p = 0.009), stages III–IV (p = 0.005), T1–T2 (0.008), and T3–T4 (0.004), were significantly different in the low- or high-risk subdivisions (Fig. 7).

Developing a nomogram that incorporates clinical features

Nomograms are always used to combine multiple factors to diagnose or predict disease onset or progression. In our study, age, gender, grade, and T, N, and M stages, clinical stage, and risk score were integrated to develop a nomogram to predict the survival probability of patients with HCC at 1, 3, and 5 years (Fig. 8A). The ROC curve showed that the AUC values corresponding to nomogram, risk score, age, gender, grade, clinical stage, and T, M, and N stages were 0.756, 0.645, 0.55, 0.444, 0.532, 0.707, 0.696, 0.521, and 0.514, respectively (Fig. 8B). The calibration plot demostrated that the nomogram performed well compared to an ideal model (Fig. 8C–E). The above results indicated that the nomogram has satisfactory accuracy in predicting the prognosis of patients with HCC.

Functional analyses based on the risk model

We further analyzed the functional enrichment analysis of DEGs (adjusted p-value < 0.05 and | log2(fold change) | ≥1) between high and low-risk groups. GO enrichment analysis revealed that these DEGs are mainly enriched in nuclear division, organelle fission, chromosome segregatio, mitotic nuclear division, and nuclear chromosome segregation. The KEGG top-ranked pathways include cell cycle, metabolism of xenobiotics by cytochrome P450, metabolism of xenobiotics by cytochrome P450, drug metabolism-cytochrome P450, and chemical carcinogenesis-DNA adducts. We found that the lower risk group with better prognosis was enriched in metabolism-related pathways, including drug metabolism-cytochrome P450, drug metabolism other enzymes, fatty acid metabolism, metabolism of xenobiotics of cytochrome P450, and primary bile acid biosynthesis based on GSEA analysis.

Assessment of differences in immune characteristics and immunotherapy between the high- and low-risk groups

Considering that TME is important for the progression and treatment of HCC, we analyzed the differences in immune characteristics between the high- and low-risk groups. The results showed in Figs. 10A and 10B hint that B cells, NK cells, neutrophils, and pDCs were highly expressed in the low-risk group, while aDCs, mast cells, macrophages, and Tfh were upregulated in the high-risk group. Additionally, immune-related functions, including cytolitic activity, MHC class, and type I and II interferon (IFN) responses, were statistically different between the two groups. Immune evasion is crucial for immunotherapy, and the greater the potential for immune evasion in patients with cancer, the less effective they are in receiving immunotherapy. We found significant differences between the high- and low-risk groups in TIDE scores, microsatellite instability (MSI) scores, immune dysfunction scores, and immune exclusion scores by analyzing the data from published immune checkpoint blockade trials on the web platform TIDE (Fig. 10C–F). The results of these scores indicate greater immune evasion potential in the high-risk group, suggesting that immunotherapy is more effective. Therefore, the risk model can predict the survival prognosis of patients with cancer and guide their clinical treatment decisions.

Pan-cancer analyses of GAS2L3 and HTR1D

Considering the scarce studies on the prognostic value of GAS2L3 and HTR1D in different cancers, we made a preliminary pan-cancer analysis in 33 cancers using the TCGA database, which we hope will be helpful to other investigators. The results showed that GAS2L3 expression was significantly increased in the following cancers compared to the adjacent paraneoplastic tissues: BCLA, BRCA, CESC, CHOL, COAD, ESCA, GBM, HNSC, KIRC, KIRP, LIHC, PCPG, SARC, STAD, THCA, and UCEC (Fig. 11A). Compared to other tumor tissues, GAS2L3 expression was highest in KIRC and SKCM, and lowest in LGG and PRAD (Fig. 11B). The Kaplan–Meier analysis was used to assess the relationship between GAS2L3 expression and survival time in various cancers, which revealed an increased GAS2L3 expression, predicted worse prognosis in ACC, LUAD, LGG, and MESO, and was associated with good prognosis in THYM (Fig. 11C). We further analyzed the relationship between GAS2L3 and clinical staging and age, and the results suggested that GAS2L3 was associated with clinical staging in ACC, KICH, and KIRP, and with age factors in BRCA, KIRP, LGG, PRAD, READ, STAD, THYM, and UCEC (Fig. 11D and 11E). The same approach was used for pan-cancer analysis of HTR1D, which was significantly increased in the following cancers compared to adjacent paraneoplastic tissues: BRCA, CHOL, COAD, ESCA, HNSC, KIRP, LIHC, LUAD, PAAD, READ, STAD, THCA, and UCEC, whereas downregulated in KICH and PRAD (Fig. 12A). Compared with other tumor tissues, HTR1D expression was highest in COAD and READ, and lowest in THYM and DLBC (Fig. 12B). HTR1D was associated with SARC, HMSC, LGG, LUAD, UVM, and THYM (Fig. 12C). Additionally, HTR1D was associated with clinical staging of BLCA, KIRC, LUAD, MESO, and THCA; and with age factor of ACC, BRCA, ESCA, OV, SARC, TGCT, THCA, THYM, and UCES (Fig. 12D and 12F).

HCC remains a global health threat, with estimates that the number of incidences will exceed one million per year by 2025¹⁷. Additionally, patients with HCC are often asymptomatic in the early stages; thus, the tumor is prone to recurrence and metastasis and the prognosis is poor, making the treatment of HCC clinically quite challenging^18,19. Traditional treatment is mainly based on surgery, supplemented by radiotherapy and chemotherapy; however, finding new therapeutic targets is urgently needed due to the high recurrence rate of tumors ^20,21. An increasing number of studies have shown that dysregulation of RNA methylation is closely associated with the development and progression of HCC, suggesting it as a new potential therapeutic target¹⁶.

In this study, first, we classified patients with HCC into two subtypes based on the found m7G genes and significant survival differences between C1 and C2. Then, we analyzed DEGs between subtypes from a prognostic perspective and found 140 genes associated with OS in HCC by univariate analysis (p < 0.05). Finally, after further narrowing by LASSO-cox and multivariate regression analyses, we identified a risk score consisting of 3 genes (GAS2L3, HTR1D, and NEIL3). Based on the median risk score, all patients with HCC of TCGA cohort were classified into high- and low-risk groups, and the Kaplan–Meier survival curve and time-dependent ROC curve showed that the risk score could effectively differentiate the prognosis of patients. More importantly, internal validation (training cohort: n = 186, and validation cohort: n = 184) confirmed the risk score as a stable and accurate prognostic assessment tool for patients with HCC. Additionally, a significant difference was found in the prognosis between the high- and low-risk groups, regardless of the clinical characteristics of patients with HCC. The risk score was determined as an independent prognostic factor for HCC by multifactorial analysis, and the model accurately predicted 1-, 3-, and 5-year survival times for patients between the high- and low-risk groups as evidenced by nomogram. NEIL3 and its close association with tumor progression has been adequately studied in the prognostic model built from three m7G-related genes. NEIL3 is a base excision repair enzyme that is thought to play a central role in the removal of damaged, mispaired bases, and DNA cross-linkage repair^22,23. NEIL3 expression levels were found to be positively correlated with total somatic mutation load and somatic single nucleotide mutation load in a variety of cancers and was further suggested to be caused by promoting the APOBEC3B expression, a potent mutagenic inducer²⁴. CircRNA produced by NEIL3 was shown to regulate ADAR1 expression, causing RNA editing in pancreatic adenocarcinoma cells, thus promoting tumor progression. A similar effect was found in lung adenocarcinoma, where NEIL3 was associated with poorer prognosis and larger tumor volume, which is consistent with our study and the mechanism of action of NEIL3 ^25,26. However, little research has been done on GAS2L3 and HTR1D in tumors; thus, we further investigated the prognostic value of both genes in other cancers. GAS2L3 is a member of the growth arrest-specific 2 protein family that regulates cell cycle and apoptosis²⁷. Few studies have found that GAS2L3 is associated with poor prognosis in glioma and HCC^28,29. HTR1D is an integral member of the 5-hydroxytryptaminergic system and interestingly, HTR1D promotes HCC proliferation and is associated with poorer prognosis, which is consistent with our study³⁰. However, HTR1D downregulation was found to promote HCC progression ³¹. Additionally, GAS2L3 was associated with prognosis of ACC, LUAD, MESO, and THYM; similarly, HTR1D was also associated with HNSC, LGG, LUAD, SARC, THYM, and UVM, which may provide possible directions for subsequent studies.

The signaling pathways associated with DEGs were analyzed between the high- and low-risk groups to further find the reasons for the prognostic differences between the two groups. We found that the top GO pathways included nuclear division, organelle fission, chromosome segregation, mitotic nuclear division, and nuclear chromosome segregation, and the top KEGG pathways included cell cycle, metabolism of xenobiotics by cytochrome P450, metabolism of xenobiotics by cytochrome P450, drug metabolism-cytochrome P450, and chemical carcinogenesis-DNA adducts. Considering that RNA modifications are associated with immune function, we revealed differences in immune cells and immune-related functions between the high- and low-risk groups. B cells, NK cells, neutrophils, and pDCs were found to be highly expressed in the low-risk group, whereas aDCs, mast cells, macrophages, and Tfh were upregulated in the high-risk group. These immune cells are closely associated with cancer development. DCs are the most powerful antigen-presenting cells and their function is related to maturation status. Immature DCs can generate immune tolerance by regulating T cell expression³²; while mature DCs enhance human cellular immunity³³. B cells are the core cells of humoral immunity and can directly kill tumor cells or promote cellular responses to inhibit tumor progression³⁴. Immune-related functions suggest cytolitic activity in the low-risk group, and type I and II IFN responses are more active, and an active immune response is usually associated with a good HCC prognosis³⁵.

In a world, our study presents the first novel m7G-based prognostic model between different subtypes of patients with HCC. The high- and low-risk groups significantly differed not only in prognosis but also between tumor microenvironment, immune escape, molecular pathways, and clinical characteristics. Our study may provide some suggestions for individualized treatment of patients with HCC and evaluate the effectiveness of prognostic immunotherapy. Additionally, it may guide the development of targeted therapeutic regimens for different subgroups of patients, thus maximizing the efficacy.

In conclusion, our model using m7G-related genes can accurately and effectively predict the prognosis of HCC patients and may provide targets for subsequent treatment.

HCC Hepatocellular carcinoma

m7G N 7-methylguanosine

RNA Ribonucleic acid

TCGA The Cancer Genome Atlas

DEGs Differential expression genes

GAS2L3 Growth arrest-specific 2-like 3

HTR1D 5-hydroxytryptamine receptor 1D

NEIL3 Nei like DNA glycosylase 3

LASSO Least absolute shrinkage and selection operator

ROC Receiver operating characteristic

AUC Area under the curve

GO Gene ontology

KEGG Kyoto Encyclopedia of Genes and Genomes

TIDE Tumor immune dysfunction and exclusion

OS Overall survival

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Availability of data and materials

All data and materials used in this study are available in The Cancer Genome Atlas (TCGA: https://cancergenome.nih.gov/).

Competing interests

The authors declare that they have no competing interests.

Funding

No funding was received.

Authors’ contributions

All authors made contributions to conception and design. JW, ZM and LY designed the study. PW, KD, YJ and QL collected data from the TCGA and performed the bioinformatics. ZM and LY drafted the main manuscript. All authors read and approved the final manuscript.

Acknowledgements

Not applicable.

Sung H, Ferlay J, Siegel RL, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA: a cancer journal for clinicians.2021;71(3):209-249.
Yang JD, Hainaut P, Gores GJ, Amadou A, Plymoth A, Roberts LR. A global view of hepatocellular carcinoma: trends, risk, prevention and management. Nature reviews Gastroenterology & hepatology.2019;16(10):589-604.
El-Serag HB. Epidemiology of viral hepatitis and hepatocellular carcinoma. Gastroenterology.2012;142(6):1264-1273.e1261.
Li J, Zhu Y. Recent Advances in Liver Cancer Stem Cells: Non-coding RNAs, Oncogenes and Oncoproteins. Frontiers in cell and developmental biology.2020;8:548335.
Li W, Chen QF, Huang T, Wu P, Shen L, Huang ZL. Identification and Validation of a Prognostic lncRNA Signature for Hepatocellular Carcinoma. Frontiers in oncology.2020;10:780.
Xu Z, Peng B, Liang Q, et al. Construction of a Ferroptosis-Related Nine-lncRNA Signature for Predicting Prognosis and Immune Response in Hepatocellular Carcinoma. Frontiers in immunology.2021;12:719175.
Wiener D, Schwartz S. The epitranscriptome beyond m(6)A. Nature reviews Genetics.2021;22(2):119-131.
Enroth C, Poulsen LD, Iversen S, Kirpekar F, Albrechtsen A, Vinther J. Detection of internal N7-methylguanosine (m7G) RNA modifications by mutational profiling sequencing. Nucleic acids research.2019;47(20):e126.
Song B, Tang Y, Chen K, et al. m7GHub: deciphering the location, regulation and pathogenesis of internal mRNA N7-methylguanosine (m7G) sites in human. Bioinformatics (Oxford, England).2020;36(11):3528-3536.
Tomikawa C. 7-Methylguanosine Modifications in Transfer RNA (tRNA). International journal of molecular sciences.2018;19(12).
Jühling F, Mörl M, Hartmann RK, Sprinzl M, Stadler PF, Pütz J. tRNAdb 2009: compilation of tRNA sequences and tRNA genes. Nucleic acids research.2009;37(Database issue):D159-162.
Moteki S, Price D. Functional coupling of capping and transcription of mRNA. Molecular cell.2002;10(3):599-609.
Ramanathan A, Robb GB, Chan SH. mRNA capping: biological functions and applications. Nucleic acids research.2016;44(16):7511-7526.
Doi Y, Wachino JI, Arakawa Y. Aminoglycoside Resistance: The Emergence of Acquired 16S Ribosomal RNA Methyltransferases. Infectious disease clinics of North America.2016;30(2):523-537.
Zhang M, Song J, Yuan W, Zhang W, Sun Z. Roles of RNA Methylation on Tumor Immunity and Clinical Implications. Frontiers in immunology.2021;12:641507.
Xie S, Chen W, Chen K, et al. Emerging roles of RNA methylation in gastrointestinal cancers. Cancer cell international.2020;20(1):585.
Llovet JM, Kelley RK, Villanueva A, et al. Hepatocellular carcinoma. Nature reviews Disease primers.2021;7(1):6.
Forner A, Llovet JM, Bruix J. Hepatocellular carcinoma. Lancet (London, England).2012;379(9822):1245-1255.
Mendizabal M, Reddy KR. Current management of hepatocellular carcinoma. The Medical clinics of North America.2009;93(4):885-900, viii.
Huo J, Cai J, Guan G, Liu H, Wu L. A Ferroptosis and Pyroptosis Molecular Subtype-Related Signature Applicable for Prognosis and Immune Microenvironment Estimation in Hepatocellular Carcinoma. Frontiers in cell and developmental biology.2021;9:761839.
Marks EI, Yee NS. Molecular Genetics and Targeted Therapy in Hepatocellular Carcinoma. Current cancer drug targets.2016;16(1):53-70.
Yang Z, Nejad MI, Varela JG, Price NE, Wang Y, Gates KS. A role for the base excision repair enzyme NEIL3 in replication-dependent repair of interstrand DNA cross-links derived from psoralen and abasic sites. DNA repair.2017;52:1-11.
Semlow DR, Zhang J, Budzowska M, Drohat AC, Walter JC. Replication-Dependent Unhooking of DNA Interstrand Cross-Links by the NEIL3 Glycosylase. Cell.2016;167(2):498-511.e414.
Shinmura K, Kato H, Kawanishi Y, et al. Abnormal Expressions of DNA Glycosylase Genes NEIL1, NEIL2, and NEIL3 Are Associated with Somatic Mutation Loads in Human Cancer. Oxidative medicine and cellular longevity.2016;2016:1546392.
Shen P, Yang T, Chen Q, et al. CircNEIL3 regulatory loop promotes pancreatic ductal adenocarcinoma progression via miRNA sponging and A-to-I RNA-editing. Molecular cancer.2021;20(1):51.
Zhao C, Liu J, Zhou H, et al. NEIL3 may act as a potential prognostic biomarker for lung adenocarcinoma. Cancer cell international.2021;21(1):228.
Zhang N, Zhao C, Zhang X, et al. Growth arrest-specific 2 protein family: Structure and function. Cell proliferation.2021;54(1):e12934.
Zhao C, Zhang N, Cui X, et al. Integrative analysis regarding the correlation between GAS2 family genes and human glioma prognosis. Cancer medicine.2021;10(8):2826-2839.
Gao X, Zhao C, Zhang N, et al. Genetic expression and mutational profile analysis in different pathologic stages of hepatocellular carcinoma patients. BMC cancer.2021;21(1):786.
Fatima S, Shi X, Lin Z, et al. 5-Hydroxytryptamine promotes hepatocellular carcinoma proliferation by influencing β-catenin. Molecular oncology.2016;10(2):195-212.
Zhou Y, Liu X, Gao Y, et al. Paeoniflorin Affects Hepatocellular Carcinoma Progression by Inhibiting Wnt/β-Catenin Pathway through Downregulation of 5-HT1D. Current pharmaceutical biotechnology.2021;22(9):1246-1253.
Dudek AM, Martin S, Garg AD, Agostinis P. Immature, Semi-Mature, and Fully Mature Dendritic Cells: Toward a DC-Cancer Cells Interface That Augments Anticancer Immunity. Frontiers in immunology.2013;4:438.
Reis e Sousa C. Dendritic cells in a mature age. Nature reviews Immunology.2006;6(6):476-483.
Tokunaga R, Naseem M, Lo JH, et al. B cell and B cell-related pathways for novel cancer treatments. Cancer treatment reviews.2019;73:10-19.
Gao Q, Zhu H, Dong L, et al. Integrated Proteogenomic Characterization of HBV-Related Hepatocellular Carcinoma. Cell.2019;179(2):561-577.e522.

Table 1 is available in the Supplementary Files section

No competing interests reported.

table1.txt

Download PDF

Version 1

posted

You are reading this latest preprint version

N 7-methylguanosine-related gene risk signature as a potential prognostic biomarker in hepatocellular carcinoma

Status:

Version 1

Abstract

Figures

Introduction

Methods And Materials

Data collection

Consensus clustering based on m7G genes

Building a prognostic model to predict the overall survival of all patients with HCC

Validation of the risk score

Independent prognostic and clinical correlation analysis

Construction of nomogram

Functional enrichment analysis

Estimation of tumor-infiltrating immune cells and immunotherapeutic evasion with risk assessment signature

GAS2L3 and HTR1D expression and survival analysis and relationship with clinical stage in cancers

Statistical analysis

Results

Tumor classification based on the m7G genes

Constructing the prognostic model from m7G-related clustering

Verification of the m7G-related genes prognostic signature

Independent prognostic value of the prognostic model

Subgroup OS analysis with different clinicopathological features for all patients with HCC in TCGA cohort

Developing a nomogram that incorporates clinical features

Functional analyses based on the risk model

Assessment of differences in immune characteristics and immunotherapy between the high- and low-risk groups

Pan-cancer analyses of GAS2L3 and HTR1D

Discussion

Conclusion

Abbreviations

Declarations

References

Table

Additional Declarations

Supplementary Files

Status:

Version 1